Font Size: a A A

Implementation Of Landslide Early Warning Model And Data Mining Based On Country & City Geologic Hazard Database Of Fujian Province

Posted on:2011-04-18Degree:MasterType:Thesis
Country:ChinaCandidate:X ZhengFull Text:PDF
GTID:2178360302992799Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Landslide's growth, development, occurrence of the process is a complex system, affected by natural and man-made and other factors, control factors in a mature or under the combined effects of multiple factors, will be generated from static to dynamic, from the amount to quality of the deformation, thus landslide.The main factors affecting the formation of a landslide based on factors and dynamic factors. The basis of factors including terrain conditions, formation lithology, geological structure, hydrogeological conditions, water development and other vegetation. Dynamic factor that triggers earthquakes, rainfall, human engineering activities.By using data mining techniques, "Fujian Province County (city, district) Division of Geological Hazard" database for data mining, on the landslide occurrence factors were analyzed to describe the probability model and CF were selected critical infrastructure caused by slippery factor sensitivity, respectively landslide location from high to low elevation, slope gradient, soil depth, land use and bedrock lithology, the dynamic factor for the total rainfall on the 3rd.Logistic regression analysis using two methods of training landslide hazard identification model, based on factors that determine the weight in the model. Then, using discriminant model on the "storm of typhoon southeast Fujian geological disaster monitoring and warning-type model" project database steep slope probability judgments.
Keywords/Search Tags:data mining, geologic hazard, coefficient of certainty, logistic, early warning model
PDF Full Text Request
Related items